pmc logo image
Logo of brjpharmJournal URL: http://www3.interscience.wiley.com/journal/121548564/home
Journal List > Br J Pharmacol > v.154(1); May 2008

Formats:
Br J Pharmacol. 2008 May; 154(1): 82–92.
Published online 2008 March 3. doi: 10.1038/bjp.2008.56.
PMCID: PMC2438981
Copyright 2008, Nature Publishing Group
Short-term or long-term treatments with a phosphodiesterase-4 (PDE4) inhibitor result in opposing agonist-induced Ca2+ responses in endothelial cells
M Campos-Toimil,1,3 T Keravis,2,3 F Orallo,1 K Takeda,2 and C Lugnier2*
1Departmento de Farmacoloxía, Facultade de Farmacia, Universidade de Santiago de Compostela, Santiago de Compostela, Spain
2Département Pharmacologie et Physico-Chimie, Université Louis Pasteur—Strasbourg I, CNRS UMR 7175, Illkirch, France
*Author for correspondence: Email: claire.lugnier/at/pharma.u-strasbg.fr
3These authors contributed equally to this work.
Received November 8, 2007; Revised January 15, 2008; Accepted January 30, 2008.
Abstract
Background and purpose: We previously reported that agonist-induced rises in cytoplasmic Ca2+ concentration ([Ca2+]i) in human umbilical vein endothelial cells (HUVEC) were inhibited after a short-term (2 min) pre-treatment with cAMP-elevating agents. The aim of this work was to study the effects of longer term (8 h) pre-treatment with dibutyryl-cAMP (db-cAMP) or rolipram, a specific inhibitor of phosphodiesterase-4 (PDE4), on [Ca2+]i, cAMP levels and PDE activity and expression in HUVEC.
Experimental approach: [Ca2+]i changes were measured in isolated HUVEC by Fura-2 imaging. Intracellular cAMP levels and PDE4 activity were assessed by enzyme-immunoassay and radio-enzymatic assay, respectively. PDE expression was measured by northern and western blot analysis.
Key results: Long-term pre-treatment of HUVEC with rolipram or db-cAMP significantly increased ATP-, histamine- and thrombin-induced [Ca2+]i rises. Short-term pre-treatment with rolipram was associated with an increase in cAMP, whereas long-term pre-treatment was associated with a decrease in cAMP. Long-term pre-treatment with rolipram or db-cAMP induced a significant increase in PDE4 activity and the expression of 74 kDa-PDE4A and 73 kDa-PDE4B was specifically enhanced. All these effects were suppressed by cycloheximide.
Conclusions and implications: Our data suggest that sustained inhibition of PDE4 by rolipram induced an increase in PDE4 activity, possibly as a compensatory mechanism to accelerate cAMP degradation and that PDE4A and PDE4B were implicated in the regulation of [Ca2+]i. Thus, isozyme-specific PDE4 inhibitors might be useful as therapeutic agents in diseases where [Ca2+]i handling is altered, such as atherosclerosis, hypertension and tolerance to β-adrenoceptor agonists.
Keywords: calcium, cAMP, rolipram, PDE4, endothelial, human primary cell
Introduction
Changes in cytoplasmic Ca2+ concentration in endothelial cells are generally accepted as key signalling events for the regulation of many endothelium-dependent processes, including the synthesis and release of vasoactive factors controlling vascular tone and endothelial permeability. Ca2+ handling in endothelial cells is modulated by intracellular cAMP levels, although the reported effects are contradictory and the underlying mechanisms are still unclear (Bolz and Pohl, 1997; Vischer and Wollheim, 1998; Hippenstiel et al., 2002). We previously reported that agonist-induced [Ca2+]i rises in human umbilical vein endothelial cells (HUVEC) were significantly reduced after 2 min pre-incubation with rolipram, a specific phosphodiesterase-4 (PDE4) inhibitor, mainly via inhibition of Ca2+ mobilization from intracellular stores (Campos-Toimil et al., 2000).
Intracellular cAMP levels are regulated by eight families of cyclic nucleotide phosphodiesterases (Bender and Beavo, 2006; Lugnier, 2006), which are differently distributed in various endothelial cells (Keravis et al., 2007). HUVEC express PDE2, PDE3 and PDE4 isozymes, with half of the cAMP-PDE activity being due to PDE4 (Favot et al., 2003). In vascular smooth muscle cells, forskolin or 8-Br-cAMP treatment for long periods induces compensatory increases in cAMP-PDE activity (Rose et al., 1997; Maurice, 1998). This increase in cAMP-PDE activity could be related to PKA-dependent phosphorylation which increases PDE4 activity (Sette et al., 1994; Liu and Maurice, 1999) and/or to enhanced expression of two PDE4D variants (Liu et al., 2000). Similarly, upregulation of PDE4 activity occurs in human myometrial cells after long-term treatment with cAMP-elevating agents (Mehats et al., 1999). To our knowledge, no data dealing with long-term cAMP-dependent regulation of [Ca2+]i and PDE in endothelial cells have been reported.
Therefore, we sought to characterize the long-term effects of cAMP-elevating agents on [Ca2+]i handling in HUVEC and to determine if they would differ from those we previously described for a short-term (2 min) treatment (Campos-Toimil et al., 2000). HUVEC were thus pre-treated for 8 h with rolipram, a selective PDE4 inhibitor (Schwabe et al., 1976; Lugnier et al., 1983, 1986) or dibutyryl-cAMP (db-cAMP), a lipophilic non-hydrolyzable cAMP analogue, which directly activates PKA. Resting and agonist-induced rises in [Ca2+]i were assessed in single HUVEC loaded with Fura-2. Total cAMP levels, cAMP-PDE and PDE4 activities from cells extracts were also determined, and PDE4 isozyme expression was analysed by northern and western blots.
Methods
Cell culture
Freshly delivered umbilical cords were obtained from a nearby hospital. HUVEC were isolated using 0.1% collagenase and grown in medium 199/RPMI 1640 (1:1, v:v) containing HEPES (10 mM), L-glutamine (2 mM), antibiotics/antimycotic (100 IU ml−1 penicillin, 100 μg ml−1 streptomycin, 0.25 μg ml−1 amphotericin B) and 20% (v:v) human serum (Klein-Soyer et al., 1986). Cells were stored in liquid N2 at the first passage and used thereafter at third to fifth passages, because cellular responses were stable for these passages.
Measurement of [Ca2+]i
HUVEC were subcultured in 75 cm2 flasks and for experiments in 35 mm Petri dishes in which a 20 mm diameter hole had been cut in the base and replaced by a thin (0.07 mm) glass coverslip. Plates were treated with polylysine (0.5 mg ml−1), sterilized by UV light and then incubated with type I collagen (0.06 mg ml−1) from rat tail at 4 °C overnight. Cells were seeded at low density (2 × 103 cells cm−2), and kept in culture (37 °C, 5% CO2 in air) for 2–4 days. Rolipram (50 μM), db-cAMP (100 μM) and cycloheximide (100 μM) were added for 8 h and then washed out. Ca2+-imaging experiments were carried out on single cells as described previously (Lynch et al., 1994; Campos-Toimil et al., 2000). Agonists (10 μM histamine, 10 μM ATP, 5000 IU l−1 thrombin) were locally applied by pressure ejection from a glass pipette (2 μm i.d.) placed 200–300 μm from target cells.
cAMP measurements
For cAMP determination, HUVEC were seeded on 96-well plates (50 000 cells per well) for 24 h (37 °C, 5% CO2 in air). Cells were incubated with rolipram (50 μM), db-cAMP (100 μM), vehicle (controls) and/or cycloheximide (100 μM) for 2 min or 8 h and then washed out. Forskolin (100 μM; 2 min) was used as a positive control (data not shown). Cells were lysed using dodecyltrimethylammonium bromide (0.5% w:v) with continuous shaking for 10 min. Total cAMP of 100 μl aliquots from each non-acetylated sample was determined by enzyme-immunoassay (cAMP Biotrak kit) following the manufacturer's instructions and evaluated using a multi-well plate reader (Titertek Multiskon® PLUS MKII, Titertek, Huntsville, AL, USA).
PDE activity, northern and western blots
Confluent HUVEC at passage 3 in 75 cm2 flasks were incubated for 8 h with rolipram (50 μM), db-cAMP (100 μM), rolipram (50 μM)+cycloheximide (100 μM), or vehicle (controls) in culture medium. Cells were washed twice with phosphate-buffered saline (in mM: 0.9 CaCl2, 0.5 MgCl2, 137 NaCl, 2.6 KCl, 10 Na2HPO4, 1.7 KH2PO4, 5.0 glucose, pH 7.3) and then harvested by scraping. Cell suspensions were centrifuged (170 g, 5 min) and pellets were stored at −80 °C.
To assay PDE activity, pellets were homogenized in buffer (in mM: 250 saccharose, 5 EGTA, 2 Mg acetate, 20 Tris, 1 dithiothreitol, 0.33 Pefabloc, 2.106 IU l−1 aprotinin, 10 mg l−1 soybean trypsin inhibitor, 10 mg l−1 leupeptin, 0.2 g l−1 BSA; pH 7.5) with an Ultra-Turrax homogenizer (3 × 10 s at 4 °C) and stored at −80 °C. Protein concentration was evaluated (Lowry et al., 1951). PDE activity was determined by radioenzymatic assay using [3H]-cAMP (Keravis et al., 1980, 2005). Total cAMP-PDE activity was assessed at 1 μM cAMP in presence of 1 mM EGTA; PDE4 activity (cAMP-PDE sensitive to rolipram) was assessed by including 50 μM rolipram in the assay.
For western blots, pellets were homogenized at 4 °C with a glass–glass Potter homogenizer (2 × 15 strokes) in buffer (in mM: 250 NaCl, 25 Tris, 5 EDTA, 1 Pefabloc, 2 mg l−1 aprotinin, 2 mg l−1 leupeptin, 2 mg l−1 pepstatin A, 1% SDS; pH 7.5). Homogenates were centrifuged (14 000 g, 2 × 10 min, 4 °C), and supernatants were aliquoted and stored at −80 °C. Protein samples (10 μg) were denatured and solubilized for 5 min at 95 °C in Laemmli buffer, electrophoresed on SDS-8% polyacrylamide gel and electro-transferred onto polyvinylidene fluoride membranes (Favot et al., 2004). Membranes were immunoblotted with anti-PDE4A (1:2000), anti-PDE4B (1:2000), anti-PDE4C (1:1000) and anti-PDE4D (1:2000) antibodies. Immobilized antigens were detected by chemiluminescence using horse radish peroxidase-conjugates as secondary antibodies (1:60 000), an ECL kit and autoradiography films. The amounts of protein in each lane were checked by re-probing the membranes with an antibody directed against glyceraldehyde-3-phosphate dehydrogenase (GAPDH), a housekeeping protein (anti-GAPDH; 1/60 000). Autoradiography signals were captured on a GeneGenius Bio Imaging System (Syngene) using the GeneSnap soft wear and analysed using the GeneTool software. Relative values for PDE signals in treated HUVEC were calculated as a percentage of untreated HUVEC, corrected for the density of GAPDH bands. GAPDH was shown not to be affected by the treatments.
For northern blots, total RNA was extracted from pellets using phenol-chloroform-guanidium thiocyanate (Chomczynski and Sacchi, 1987), quantified at 260 nm and stored at −80 °C with two volumes of formamide as 20 μg aliquots. RNA was size-fractionated by agarose-formaldehyde gel electrophoresis (Lehrach et al., 1977), transferred to nylon membranes by upward capillary transfer and immobilized by UV crosslinking. Membranes were hybridized at 68 °C in Denhardt's reagent using cDNA probes for PDE4A, PDE4B, PDE4C and PDE4D radiolabeled with [α-32P]-dCTP by random priming (Prime-It II kit) and then washed twice in 20 × saline Na citrate (SSC) in presence of 0.1% SDS at 20 °C for 2 × 10 min (20 × SSC contained 3 M NaCl, 0.3 M Na citrate, pH 7.0). Membranes were subjected to increasing stringencies for 2 × 15 min (0.2 × SSC at 20 °C and 42 °C, 0.1 × SSC at 42 °C) followed by autoradiography (Hyperfilms MP). RNA in each lane was quantified by monitoring 18S and 28S ribosomal RNA on ethidium bromide-stained agarose gels.
Data presentation and statistical analysis
Ca2+ data are expressed as mean±s.e.mean of at least 18 individual cells. The area under the Ca2+ curve obtained from individual cells was determined by the trapezoid rule (Prism 2 software; Graphpad): the curve is divided into series of trapezoids, and then the area of each trapezoid is calculated individually (Burden and Faires, 2005). cAMP data are expressed as % of basal cAMP value (39.4±7.2 pmol 10−6 cells) and are mean±s.e.mean of three independent experiments. PDE activities are expressed in pmol min−1 mg−1 as mean±s.d. of two independent experiments. Northern and western blots were evaluated using a GeneGenius Bio Imaging System (Syngene) and GeneSnap and GeneTools software. Relative values for western blot PDE signals in treated HUVEC were calculated as % of untreated HUVEC, corrected for GAPDH density. Results are expressed as mean±s.e.mean of three independent experiments.
Student's 2-tailed t-test for unpaired data or one-way ANOVA (followed by a Bonferroni's post hoc test when appropriate) were used for statistical analysis, with P<0.05 being considered significant.
Drugs and chemicals
[3H]-cAMP (34 Ci mmol−1), [α-32P]-dCTP (3000 Ci mmol−1), dodecyltrimethylammonium bromide, ECL kits, Hybond-P polyvinylidene fluoride membrane, Hyperfilms MP and Kodak Biomax films were from Amersham, Little Chalfont, England. Aprotinin, leupeptin, pepstatin and 4-(2-aminoethyl)-benzenesulphonyl fluoride HCl (Pefabloc) were from Interchim. Agarose, L-glutamine, medium 199 and RPMI 1640 were from Gibco, Cergy Pontoise, France. Endotoxin-free human serum was from ‘Etablissement de Transfusion Sanguine' (Strasbourg, France; the serum was obtained from a pool of 13–15 healthy donors negative for hepatitis B virus and HIV and it was complement-inactivated at 56 °C for 30 min). Collagenase B was from Roche Diagnostics, Meylan, France. Horse radish peroxidase-conjugated secondary antibodies were from Promega, Charbonnières-les-Bains, France. Anti-GAPDH and anti-PDE4C antibodies were respectively from FabGennix, Frisco, Texas, USA. Acrylamide/bis-acrylamide (29:1 mix ratio; 30% solution), amphotericin B, ATP, BSA, chloroform, cycloheximide, db-cAMP, forskolin, histamine, penicillin, poly-L-lysine hydrobromide and streptomycin were from Sigma, l'Isle d'Abeau Chenes, Saint Quentin Fallavier, France. Prime-It II kit was from Stratagene, Amsterdam, The Netherlands. Denhardt's reagent, formaldehyde, guanidium thiocyanate, Tris, NaCl, nylon membrane, phenol and SDS were from Q.BIOgene. All other reagents were analytical grade (Merck, Darmstadt, Germany).
Rolipram and human α-thrombin were gifts from Schering AG (Berlin, Germany) and Dr JM Freyssinet (INSERM U143, Strasbourg), respectively. Anti-PDE4A (AC55), anti-PDE4B (K118) and anti-PDE4D (M3S1) (Jin et al., 1998) antibodies were gifts of Dr Marco Conti (Stanford University). Rat PDE4-cDNAs were also obtained from Dr Conti: testis 1.0 kb PDE4A-cDNA, Sertoli cell 2.4 kb PDE4B-cDNA, testis 2.3 kb PDE4C-cDNA and Sertoli cell 1.9 kb PDE4D-cDNA (Swinnen et al., 1989).
All drugs used were dissolved in water, except for rolipram, forskolin and cycloheximide, which were dissolved in DMSO (final concentration [less-than-or-eq, slant]0.2%). Rolipram was used also to assess PDE4 activity and DMSO was present in the assay at a final concentration of 1%. Controls were treated with solvent alone.
Results
Effects of rolipram and db-cAMP on agonist-induced [Ca2+]i rises
In normal external solution (containing 2 mM Ca2+), the basal [Ca2+]i in control HUVEC was 86. 6±0.5 nM (n=84) and was unchanged over the experimental time course. After 8 h pre-treatment with rolipram (50 μM) or db-cAMP (100 μM), basal [Ca2+]i was unchanged.
In control cells, local applications of agonist (histamine, ATP, thrombin) for 60 s, caused rapid increases in [Ca2+]i accompanied by sustained plateau-like elevations of varying duration (Figure 1Figure 1). Return to near basal levels occurred within 120–140 s for histamine (Figure 1aFigure 1) and ATP (Figure 1bFigure 1), and within 180–200 s for thrombin (Figure 1cFigure 1). The percentage of responsive cells in a given culture did not vary significantly and was 65–70% for ATP and 90–95% for histamine and thrombin. After 8 h pre-treatment with 50 μM rolipram, [Ca2+]i increases induced by the same agonists were significantly potentiated (Figure 1Figure 1, left panels). In a similar fashion, agonist-induced increases in [Ca2+]i were significantly potentiated after 8 h pre-treatment with 100 μM db-cAMP (Figure 1Figure 1, right panels). Note that the amplitudes of both the initial peak and the plateau phase were increased. The potentiation of histamine-induced [Ca2+]i responses caused by 8 h pre-treatment with rolipram and db-cAMP was abolished when 100 μM cycloheximide was included during the pre-treatment (Table 1).
Figure 1
Figure 1
Figure 1
Effects of 8 h pre-treatment of HUVEC with 50 μM rolipram (left panels) or 100 μM db-cAMP (right panels) on [Ca2+]I increases induced by local applications of 10 μM histamine (a), 10 μ (more ...)
Table 1
Table 1
Cycloheximide reverses effects of pre-treatment with rolipram or db-cAMP on histamine-induced [Ca2+]i increases in HUVEC, bathed in normal Ca2+ solution
Effects of rolipram and db-cAMP on histamine-induced [Ca2+]i rises in Ca2+-free solution
In absence of external Ca2+, histamine application evoked a rapid elevation in [Ca2+]i that returned to basal values within 80–90 s. The initial peak of the histamine-induced response was similar to that observed in the presence of 2 mM Ca2+, whereas the plateau phase was completely suppressed (Figure 2aFigure 2). Thus, as commonly held, initial [Ca2+]i responses are largely due to release of Ca2+ from intracellular stores while the sustained plateau phase requires Ca2+ entry (Bregestovski et al., 1988; Duchêne and Takeda, 1997). The potentiation of the histamine-induced Ca2+ response after 8 h pre-treatment with rolipram (Figure 2bFigure 2) or db-cAMP (Figure 2cFigure 2) was preserved in the absence of external Ca2+.
Figure 2
Figure 2
Figure 2
(a) [Ca2+]i increases produced by local application of 10 μM histamine in Ca2+-free external solution compared with 2 mM Ca2+-containing external solution. Data are mean±s.e.mean from at least 18 (more ...)
Effects of rolipram and db-cAMP on intracellular cAMP levels
The basal cAMP level measured immediately in HUVEC was 39.4±7.2 pmol 10−6 cells (n=3) and was unchanged after 8 h. As shown in Figure 3Figure 3, a 2 min or 8 h pre-treatment with 100 μM cycloheximide caused no significant changes in basal cAMP levels. A 2 min pre-incubation of HUVEC with 50 μM rolipram led to a 84% increase in cAMP, that was not affected by cycloheximide (Figure 3aFigure 3). In contrast, an 8 h pre-treatment with rolipram provoked a significant decrease (62%) of cAMP levels and the presence of cycloheximide was able to restore the basal level (Figure 3aFigure 3). On the other hand, both 2 min and 8 h pre-treatments with 100 μM db-cAMP increased cAMP levels (209 and 188%, respectively) and cycloheximide was without any significant effect (Figure 3bFigure 3).
Figure 3
Figure 3
Figure 3
Intracellular cAMP levels. HUVEC were pre-treated for 2 min or 8 h with 50 μM rolipram (a) or 100 μM db-cAMP (b), in presence or absence of 100 μM cycloheximide. Data are mean±s.e.mean (more ...)
Effects of rolipram and db-cAMP on cAMP-PDE activity
PDE activity was mainly directed towards cAMP in control HUVEC, with cAMP-PDE activity being seven-fold higher than that of cGMP-PDE (respectively, 66.8±1.5 vs 10.0±0.5 pmol min−1 mg−1). An 8 h pre-treatment of HUVEC with 50 μM rolipram or 100 μM db-cAMP increased total cAMP-PDE activity in cell extracts and the effect of rolipram was abolished by cycloheximide (Figure 4aFigure 4). The PDE4 activity of control extracts (27.6±2.9 pmol min−1 mg−1) represents 41% of total cAMP-PDE activity (66.8±1.4 pmol min−1 mg−1). An 8 h pre-treatment of HUVEC with 50 μM rolipram or 100 μM db-cAMP increased the PDE4 activity of extracts and cycloheximide blocked the effect of rolipram (Figure 4bFigure 4).
Figure 4
Figure 4
Figure 4
Effects of rolipram and db-cAMP pre-treatment (8 h) on cAMP-PDE activity, in presence or absence of 100 μM cycloheximide. Total cAMP-PDE activity (a) and PDE4 activity (b) were assessed as described in Methods. Data are mean±s.d. (more ...)
After rolipram pre-treatment, the increase in PDE4 activity (+134%; Figure 4bFigure 4) was two-fold higher than the increase in total cAMP-PDE activity (+62%; Figure 4aFigure 4), indicating that an increase in PDE4 activity was mainly involved. After db-cAMP pre-treatment, the total cAMP-PDE and PDE4 activities were increased similarly (+54%; Figures 4a and bFigure 4). Cycloheximide suppressed the increase in PDE4 activity due to rolipram (Figure 4bFigure 4), suggesting that this increase could be due to increased synthesis of PDE4. Thus, we investigated the effects of these pre-treatments on PDE4 expression.
Effects of rolipram and db-cAMP on PDE4 isozyme expression
The PDE4 isozyme family comprises four subtypes: PDE4A, PDE4B, PDE4C and PDE4D. Western blot analysis was performed to investigate whether a specific isozyme was responsible for the increased PDE4 activity. The PDE4 isozyme expression was as follows: a single signal of 74 kDa for PDE4A; three signals of 102, 89 and 73 kDa respectively in a ratio of 1, 1.5 and 0.9 for PDE4B; a single signal of 74 kDa for PDE4C; a single signal of 64 kDa for PDE4D.
On western blots, PDE4A-protein was expressed as a single band at 74 kDa (Figure 5aFigure 5). The amounts of PDE4A were significantly increased by pre-treatment with db-cAMP and rolipram, with cycloheximide reversing the rolipram-induced increase. Northern blot analysis (Figure 5bFigure 5) shows that PDE4A-mRNA was expressed as a 3.7 kb signal which was increased by pre-treatments with db-cAMP (+133%) and rolipram (+74%). Cycloheximide abolished the rolipram increase of the PDE4A transcript.
Figure 5
Figure 5
Figure 5
Effects of rolipram and db-cAMP pre-treatment (8 h) on expression of PDE4A. (a) Amounts of PDE4A protein were determined by western blot using the AC55 anti-PDE4A antibody, as illustrated by a representative gel, followed by densitometric analysis. (more ...)
PDE4B-protein was expressed as three bands of 102, 89 and 73 kDa (Figures 6a–cFigure 6). Pre-treatments with db-cAMP and rolipram induced no change in the 102 and 89 kDa bands. However, the 73 kDa band was significantly increased by db-cAMP and rolipram with the latter increase being abolished by cycloheximide. Northern blot analysis shows that PDE4B-mRNA was expressed as a 3.7 kb signal which was increased by db-cAMP (+95%) and rolipram (+24%); cycloheximide abolished the rolipram-induced increase of the PDE4B transcript (Figure 6dFigure 6).
Figure 6
Figure 6
Figure 6
Effects of rolipram and db-cAMP pre-treatment (8 h) on expression of PDE4B. (a–c) Amounts of PDE4B-protein were determined by western blot using the K118 anti-PDE4B antibody, as illustrated by representative gels, followed by densitometric (more ...)
PDE4C was expressed as a 74 kDa protein and a 3.7 kb transcript. The amounts of protein and mRNA were not significantly modified by db-cAMP or rolipram pre-treatments (Figure 7Figure 7).
Figure 7
Figure 7
Figure 7
Effects of rolipram and db-cAMP pre-treatment (8 h) on expression of PDE4C. (a) Amounts of PDE4C-protein were determined by western blot using the FabGennix anti-PDE4C antibody, as illustrated by a representative gel, followed by densitometric (more ...)
PDE4D-protein (Figure 8aFigure 8) was expressed as a single band at 64 kDa, which was significantly increased by db-cAMP and rolipram pre-treatment. In this case, including cycloheximide together with rolipram during pre-treatment still resulted in increased PDE4D expression (+27%). Northern blot analysis (Figure 8bFigure 8) shows that the 3.9 kb PDE4D-mRNA was increased by db-cAMP (+58%), rolipram (+32%) and rolipram+cycloheximide (+25%) pretreatment. Again, cycloheximide did not reverse the rolipram-induced increase of the 3.9 kb PDE4D-transcript. A 7.8 kb transcript for PDE4D was also detected and was not modified by either db-cAMP or rolipram. However, cycloheximide+rolipram increased the expression of this transcript by 139% (Figure 8bFigure 8).
Figure 8
Figure 8
Figure 8
Effects of rolipram and db-cAMP pre-treatment (8 h) on expression of PDE4D. (a) Amounts of PDE4D-protein were revealed by western blot using the M3S1 anti-PDE4D antibody. Results are expressed as % of control values (mean±s.e.mean (more ...)
Discussion
In the present study, we demonstrate that 8 h pre-treatment of HUVEC with rolipram (a specific PDE4 inhibitor) or db-cAMP (a permeant non-hydrolyzable cAMP-analogue) significantly increased agonist-induced [Ca2+]i rises. Including cycloheximide during pre-treatment blocked the potentiation of histamine-induced [Ca2+]i increases, indicating that protein synthesis was necessarily involved. The results presented here clearly contrast with our previous report on the effects of short term (2 min) pre-treatment of HUVEC with the same agents, which resulted in a significant decrease of agonist-induced [Ca2+]i rises (Campos-Toimil et al., 2000).
We found that the effects of rolipram on intracellular cAMP levels in HUVEC also differ depending on the time of pre-treatment. cAMP levels were significantly increased by 2 min pre-treatment with rolipram, in good agreement with results obtained using IBMX, a non-selective PDE inhibitor (Hopkins and Gorman, 1981) and RP73401, a selective PDE4 inhibitor (Keravis et al., 2007). This indicates that HUVEC have significant basal AC activity. In contrast, 8 h pre-treatment with rolipram surprisingly and significantly decreased cAMP levels. This last effect was abolished in the presence of cycloheximide, consistent with de novo PDE4 protein synthesis being involved in the rolipram-induced decrease in cAMP. As expected, a 2 min pre-treatment with db-cAMP significantly increased cAMP level. After 8 h pre-treatment of HUVEC with db-cAMP, cAMP levels were also markedly increased, indicating that the non-hydrolyzable db-cAMP is recognized as cAMP by enzyme-immunoassay, as reported previously (Hosokawa et al., 2001). Although db-cAMP is able to stimulate PDE4 expression, as it is not hydrolysed, its effect on intracellular cAMP level was not time-dependent. Therefore, the presence of cycloheximide did not modify the increased cAMP level.
Among the four PDE4 subtypes (PDE4A, PDE4B, PDE4C and PDE4D) present in HUVEC, the expression of 74 kDa-PDE4A and 73 kDa-PDE4B was clearly increased after sustained inhibition of PDE4 activity by 8 h pre-treatment with rolipram and these increases were reversed by cycloheximide. The effect of rolipram on PDE4 expression was mimicked by db-cAMP pre-treatment. The increased PDE4A and PDE4B expression is likely to be mediated by the sustained elevation in cAMP, as previously described for PDE4D in smooth muscle cells (Rose et al., 1997; Maurice, 1998; Tilley and Maurice, 2002). Similarly, PDE4A and PDE4B were also induced in rat pulmonary microvascular endothelial cells, following long-term treatment with forskolin and rolipram, and this induction was modulated by intracellular cAMP content (Zhu et al., 2004).
Here, the increased expression of PDE4A and PDE4B after sustained pre-treatment was accompanied by an increase in total cAMP-PDE activity, due largely to an increase in PDE4 activity. However, rolipram pre-treatment induced a significantly greater contribution of PDE4 to the increase in total cAMP-PDE activity compared with db-cAMP. This may arise from a possible localized regulation of PDE4 as well as a non-specific effect of db-cAMP on PDE activity.
Interestingly, the expression of 64 kDa-PDE4D was regulated differently compared with the other PDE4 isozymes. As 64 kDa-PDE4D expression was increased by rolipram and db-cAMP pre-treatment, cycloheximide did not reverse the effect of rolipram. This suggests that PDE4D might be down-regulated by a repressor such as the inducible cAMP early repressor (Lamas and Sassone-Corsi, 1997) and consequently, differently involved in PDE activity and resulting cAMP levels. Indeed, it was shown that PDE4D is localized in the perinuclear region (Jin et al., 1998) and is associated with the nuclear envelope (Lugnier et al., 1999). This specific localization of the PDE4D subtype may regulate cAMP dependent transcription factors in the vicinity of the nuclear envelope.
Taken together, our results indicate that a sustained elevation of intracellular cAMP after long term pre-treatment with rolipram or db-cAMP leads to increased expression and activity of PDE4, probably as a compensatory mechanism to accelerate cAMP degradation. We propose that this effect explains, at least in part, the differences in agonist-induced Ca2+ responses observed after short (Campos-Toimil et al., 2000) and long term pre-treatments with cAMP-elevating agents.
It is generally accepted that agonists like histamine, ATP and thrombin induce rises in [Ca2+]i by activation of specific G-protein-coupled receptors, inositol 1,4,5-trisphosphate- (IP3) mediated release of Ca2+ from internal stores and Ca2+ influx via several types of Ca2+-permeable channels, including store-operated Ca2+ channels responsible for capacitative Ca2+ entry and agonist-activated nonselective cationic channels (Bregestovski et al., 1988; Carter et al., 1988; Hallam et al., 1988; von der Weid et al., 1993; Duchêne and Takeda, 1997; Nilius and Droogmans, 2001; Tiruppathi et al., 2006). In agreement, our results indicate that the initial transient histamine-induced elevation in [Ca2+]i is due to mobilization of Ca2+ from intracellular stores, while the sustained plateau requires Ca2+ influx, as previously described in HUVEC (Oike et al., 1994; Lantoine et al., 1998; Campos-Toimil et al., 2000).
The interaction of cAMP with Ca2+ signals in HUVEC appears to occur during the mobilization of intracellular Ca2+ as the potentiation of histamine-induced rises in [Ca2+]i after 8 h pre-treatment with rolipram or db-cAMP was also observed in the absence of extracellular Ca2+. Moreover, in our previous short-term pre-treatment study (Campos-Toimil et al., 2000), reduction of agonist-induced rises in [Ca2+]i by cAMP-elevating agents was mainly due to inhibition of Ca2+ mobilization from internal stores, suggesting that changes in cAMP levels may directly regulate release of intracellular Ca2+. Finally, despite the existence of contradictory data, most studies indicate that elevated cAMP inhibits phosphoinositide hydrolysis and IP3 generation in a variety of cell types (Socorro et al., 1990; Keularts et al., 2000; for review, see Bruce et al., 2003). Furthermore, inhibitors of cAMP-PDE decrease IP3 generation in smooth muscle (Challiss et al., 1998) and this may involve specific PDE4 isozymes at the plasma membrane, as shown by cytochemistry (Okruhlicova et al., 1996). It is important to note that PDE4 isozymes may have different subcellular localizations (Jin et al., 1998; Lugnier, 2006; Houslay et al., 2007), thereby resulting in local regulation of discrete cAMP pools by specific PDE4 isozymes. The interaction between elevated cAMP and Ca2+ release could also occur by modulation of IP3 receptors, as cAMP alters the characteristics of these receptors via PKA-dependent phosphorylation (Bugrim, 1999). This latter effect might involve a PDE4 isoform associated with the sarcoplasmic reticulum (Lugnier et al., 1993).
The reduced release of intracellular Ca2+ by elevated cAMP level may indirectly decrease the subsequent entry of Ca2+ via store-operated Ca2+ channels. This would represent another mechanism for cAMP regulation of endothelial function, as capacitative Ca2+ entry in HUVEC is crucial in controlling endothelial permeability and the release of endothelial factors, including NO (Lin et al., 2000; Nilius and Droogmans, 2001; Tiruppathi et al., 2006). It was also recently shown that an increase in Ca2+ caused relocalization of PDE4A1 from trans-Golgi to punctate structures, which would allow further possible crosstalk between cAMP and Ca2+ signalling pathways (Huston et al., 2006).
In view of the above considerations, our results and the derived mechanistic hypotheses are summarized in Figure 9Figure 9. A brief elevation of cAMP due to short-term (2 min) inhibition of PDE4 with rolipram or directly due to addition of db-cAMP reduces agonist-induced Ca2+ rises, by decreasing Ca2+ release from internal stores, although Ca2+ influx may also be altered. This agrees with a decrease in histamine-induced Ca2+ signals by cAMP-elevating agents described in HUVEC (Bolz and Pohl, 1997), even if others found no significant effects of such agents on agonist-evoked rises in [Ca2+]i (Carson et al., 1989; Vischer and Wollheim, 1998). A sustained elevation of cAMP induced by 8 h pre-treatment with rolipram leads to increased expression and activity of PDE4, resulting finally in a marked decrease of cAMP. This might result in an increase of 5′-AMP which would participate in activation of AMP-kinase (see McGee and Hargreaves, 2008). In these conditions, agonist-induced increases in [Ca2+]i are potentiated as the inhibitory effect of cAMP, no longer exists. On the other hand, if a PDE4 inhibitor was permanently present, specific inhibition of the induced PDE4 isozymes might induce a decrease of [Ca2+]i.
Figure 9
Figure 9
Figure 9
Proposed scheme accounting for the short- and long-term effects of rolipram pre-treatment in HUVEC. Agonists induce rises in [Ca2+]i and may stimulate AC activity. After 2 min pre-treatment with cAMP-elevating agents (rolipram, db-cAMP), (more ...)
In conclusion, we show for the first time that long-term pre-treatment with cAMP-elevating agents significantly increases agonist-induced [Ca2+]i rises in HUVEC, probably via enhanced Ca2+ mobilization from intracellular stores. We suggest that this effect involves cAMP-dependent upregulation of PDE4 expression, in particular PDE4A and PDE4B. Consequently, PDE4 isozymes may represent useful targets for therapeutic intervention. The development of isozyme-specific PDE4 inhibitors could be helpful as selective therapeutic agents in vascular diseases where [Ca2+]i handling is altered, such as atherosclerosis and hypertension. Finally, an increase in PDE4 activity may be involved in the development of tachyphylaxis following the administration of the β-adrenoceptor agonists (Mehats et al., 1999). Thus, cAMP-dependent upregulation of PDE4A and PDE4B may participate in tolerance to β-adrenoceptor agonists and consequently PDE4 isotype inhibitors could represent a new and promising therapy for tolerance induced by β-adrenoceptor stimulation.
Acknowledgments
This work was supported in part by the CNRS, Inserm, the Ligue Nationale Contre le Cancer (Fondation Aprifel), the Spanish Ministerio de Ciencia y Tecnología (RYC2002-10) and the Spanish Ministerio de Sanidad y Consumo (Spain; FISS PI061537). Francisco Orallo is especially grateful to the Consellería de Educación e Ordenación Universitaria, Xunta de Galicia (Spain) for giving him financial support to increase his research activity and to reduce his teaching activity during the academic year 2007–2008 (Programa de promoción de intensificación de la actividad investigadora en el Sistema Universitario de Galicia). We are very grateful to Dr Marco Conti for antibodies and cDNAs for PDE4 subtypes. We thank Schering (Berlin) for rolipram. We acknowledge the Service Maternité du Centre Médico Chirurgical et Obstétrical de Strasbourg for providing us with umbilical cords.
Abbreviations
[Ca2+]icytoplasmic Ca2+ concentration
db-cAMPdibutyryl-cAMP
HUVEChuman umbilical vein endothelial cells

Notes
Conflict of interest
The authors state no conflict of interest.
References
  • Bender AT, Beavo JA. Cyclic nucleotide phosphodiesterases: molecular regulation to clinical use. Pharmacol Rev. 2006;58:488–520. [PubMed]
  • Bolz SS, Pohl U. Indomethacin enhances endothelial NO release--evidence for a role of PGI2 in the autocrine control of calcium-dependent autacoid production. Cardiovasc Res. 1997;36:437–444. [PubMed]
  • Bregestovski P, Bakhramov A, Danilov S, Moldobaeva A, Takeda K. Histamine-induced inward currents in cultured endothelial cells from human umbilical vein. Br J Pharmacol. 1988;95:429–436. [PubMed]
  • Bruce JI, Straub SV, Yule DI. Crosstalk between cAMP and Ca2+ signaling in non-excitable cells. Cell Calcium. 2003;34:431–444. [PubMed]
  • Bugrim AE. Regulation of Ca2+ release by cAMP-dependent protein kinase. A mechanism for agonist-specific calcium signaling. Cell Calcium. 1999;25:219–226. [PubMed]
  • Burden RL, Faires JD. Numerical differentiation and integration Numerical Analysis 2005. Thomson/Brooks Cole: Belmont, CA; 167–248.In: Burden RL, Faires JD (eds)8th edn.
  • Campos-Toimil M, Lugnier C, Droy-Lefaix MT, Takeda K. Inhibition of type 4 phosphodiesterase by rolipram and Ginkgo biloba extract (EGb 761) decreases agonist-induced rises in internal calcium in human endothelial cells. Arterioscler Thromb Vasc Biol. 2000;20:E34–E40. [PubMed]
  • Carson MR, Shasby SS, Shasby DM. Histamine and inositol phosphate accumulation in endothelium: cAMP and a G protein. Am J Physiol. 1989;257:L259–L264. [PubMed]
  • Carter TD, Hallam TJ, Cusack NJ, Pearson JD. Regulation of P2y-purinoceptor-mediated prostacyclin release from human endothelial cells by cytoplasmic calcium concentration. Br J Pharmacol. 1988;95:1181–1190. [PubMed]
  • Challiss RA, Adams D, Mistry R, Nicholson CD. Modulation of spasmogen-stimulated Ins(1,4,5)P3 generation and functional responses by selective inhibitors of types 3 and 4 phosphodiesterase in airways smooth muscle. Br J Pharmacol. 1998;124:47–54. [PubMed]
  • Chomczynski P, Sacchi N. Single-step method of RNA isolation by acid guanidinium thiocyanate- phenol-chloroform extraction. Anal Biochem. 1987;162:156–159. [PubMed]
  • Duchêne AD, Takeda K. P2Y- and P2U-mediated increases in internal calcium in single bovine aortic endothelial cells in primary culture. Endothelium. 1997;5:277–286. [PubMed]
  • Favot L, Keravis T, Holl V, Le Bec A, Lugnier C. VEGF-induced HUVEC migration and proliferation are decreased by PDE2 and PDE4 inhibitors. Thromb Haemost. 2003;90:334–343. [PubMed]
  • Favot L, Keravis T, Lugnier C. Modulation of VEGF-induced endothelial cell cycle protein expression through cyclic AMP hydrolysis by PDE2 and PDE4. Thromb Haemost. 2004;92:634–645. [PubMed]
  • Hallam TJ, Jacob R, Merritt JE. Evidence that agonists stimulate bivalent-cation influx into human endothelial cells. Biochem J. 1988;255:179–184. [PubMed]
  • Hippenstiel S, Witzenrath M, Schmeck B, Hocke A, Krisp M, Krull M, et al. Adrenomedullin reduces endothelial hyperpermeability. Circ Res. 2002;91:618–625. [PubMed]
  • Hopkins NK, Gorman RR. Regulation of endothelial cell cyclic nucleotide metabolism by prostacyclin. J Clin Invest. 1981;67:540–546. [PubMed]
  • Hosokawa M, Sato A, Ishigamori H, Kohno H, Tanaka T, Takahashi K. Synergistic effects of highly unsaturated fatty acid-containing phosphatidyl-ethanolamine on differentiation of human leukemia HL-60 cells by dibutyryl cyclic adenosine monophosphate. Jpn J Cancer Res. 2001;92:666–672. [PubMed]
  • Houslay MD, Baillie GS, Maurice DH. cAMP-Specific phosphodiesterase-4 enzymes in the cardiovascular system: a molecular toolbox for generating compartmentalized cAMP signaling. Circ Res. 2007;100:950–966. [PubMed]
  • Huston E, Gall I, Houslay TM, Houslay MD. Helix-1 of the cAMP-specific phosphodiesterase PDE4A1 regulates its phospholipase-D-dependent redistribution in response to release of Ca2+ J Cell Sci. 2006;119:3799–3810. [PubMed]
  • Jin SL, Bushnik T, Lan L, Conti M. Subcellular localization of rolipram-sensitive, cAMP-specific phosphodiesterases. Differential targeting and activation of the splicing variants derived from the PDE4D gene. J Biol Chem. 1998;273:19672–19678. [PubMed]
  • Keravis T, Silva AP, Favot L, Lugnier C. Role of PDEs in vascular health and disease: endothelial PDEs and angiogenesis Cyclic Nucleotide Phosphodiesterases in Health and Diseases 2007. CRC Press: Boca Raton, FL; 417–439.In: Beavo JA, Francis SH, Houslay MD (eds).
  • Keravis T, Thaseldar-Roumié R, Lugnier C. Assessment of phosphodiesterase isozyme contribution in cell and tissue extracts Phosphodiesterase Methods and Protocols. Methods in Molecular Biology Vol 307 2005. Humana Press: Totowa, NJ; 63–74.In: Lugnier C (ed).
  • Keravis T, Wells JN, Hardman JG. Cyclic nucleotide phosphodiesterase activities from pig coronary arteries. Lack of interconvertibility of major forms. Biochim Biophys Acta. 1980;613:116–129. [PubMed]
  • Keularts IM, van Gorp RM, Feijge MA, Vuist WM, Heemskerk JW. Alpha(2A)-adrenergic receptor stimulation potentiates calcium release in platelets by modulating cAMP levels. J Biol Chem. 2000;275:1763–1772. [PubMed]
  • Klein-Soyer C, Beretz A, Millon-Collard R, Abecassis J, Cazenave JP. A simple in vitro model of mechanical injury of confluent cultured endothelial cells to study quantitatively the repair process. Thromb Haemost. 1986;56:232–235. [PubMed]
  • Lamas M, Sassone-Corsi P. The dynamics of the transcriptional response to cyclic adenosine 3′,5′-monophosphate: recurrent inducibility and refractory phase. Mol Endocrinol. 1997;11:1415–1424. [PubMed]
  • Lantoine F, Iouzalen L, Devynck MA, Millanvoye-Van Brussel E, David-Dufilho M. Nitric oxide production in human endothelial cells stimulated by histamine requires Ca2+ influx. Biochem J. 1998;330:695–699. [PubMed]
  • Lehrach H, Diamond D, Wozney JM, Boedtker H. RNA molecular weight determinations by gel electrophoresis under denaturing conditions, a critical reexamination. Biochemistry. 1977;16:4743–4751. [PubMed]
  • Lin S, Fagan KA, Li KX, Shaul PW, Cooper DM, Rodman DM. Sustained endothelial nitric-oxide synthase activation requires capacitative Ca2+ entry. J Biol Chem. 2000;275:17979–17985. [PubMed]
  • Liu H, Maurice DH. Phosphorylation-mediated activation and translocation of the cyclic AMP- specific phosphodiesterase PDE4D3 by cyclic AMP-dependent protein kinase and mitogen-activated protein kinases. A potential mechanism allowing for the coordinated regulation of PDE4D activity and targeting. J Biol Chem. 1999;274:10557–10565. [PubMed]
  • Liu H, Palmer D, Jimmo SL, Tilley DG, Dunkerley HA, Pang SC, et al. Expression of phosphodiesterase 4D (PDE4D) is regulated by both the cyclic AMP-dependent protein kinase and mitogen-activated protein kinase signaling pathways. A potential mechanism allowing for the coordinated regulation of PDE4D activity and expression in cells. J Biol Chem. 2000;275:26615–26624. [PubMed]
  • Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein measurement with the folin phenol reagent. J Biol Chem. 1951;193:265–275. [PubMed]
  • Lugnier C. Cyclic nucleotide phosphodiesterase (PDE) superfamily: a new target for the development of specific therapeutic agents. Pharmacol Ther. 2006;109:366–398. [PubMed]
  • Lugnier C, Keravis T, Le Bec A, Pauvert O, Proteau S, Rousseau E. Characterization of cyclic nucleotide phosphodiesterase isoforms associated to isolated cardiac nuclei. Biochim Biophys Acta. 1999;1472:431–446. [PubMed]
  • Lugnier C, Muller B, Le Bec A, Beaudry C, Rousseau E. Characterization of indolidan- and rolipram-sensitive cyclic nucleotide phosphodiesterases in canine and human cardiac microsomal fractions. J Pharmacol Exp Ther. 1993;265:1142–1151. [PubMed]
  • Lugnier C, Schoeffter P, Le Bec A, Strouthou E, Stoclet JC. Selective inhibition of cyclic nucleotide phosphodiesterases of human, bovine and rat aorta. Biochem Pharmacol. 1986;35:1743–1751. [PubMed]
  • Lugnier C, Stierle A, Beretz A, Schoeffter P, Lebec A, Wermuth CG, et al. Tissue and substrate specificity of inhibition by alkoxy-aryl-lactams of platelet and arterial smooth muscle cyclic nucleotide phosphodiesterases relationship to pharmacological activity. Biochem Biophys Res Commun. 1983;113:954–959. [PubMed]
  • Lynch JW, Lemos VS, Bucher B, Stoclet JC, Takeda K. A pertussis toxin-insensitive calcium influx mediated by neuropeptide Y2 receptors in a human neuroblastoma cell line. J Biol Chem. 1994;269:8226–8233. [PubMed]
  • Maurice DH. Cyclic nucleotide-mediated regulation of vascular smooth muscle cell cyclic nucleotide phosphodiesterase activity. Selective effect of cyclic AMP. Cell Biochem Biophys. 1998;29:35–47. [PubMed]
  • McGee SL, Hargreaves M. AMPK and transcriptional regulation. Front Biosci. 2008;13:3022–3033. [PubMed]
  • Mehats C, Tanguy G, Dallot E, Robert B, Rebourcet R, Ferre F, et al. Selective up-regulation of phosphodiesterase-4 cyclic adenosine 3′,5′- monophosphate (cAMP)-specific phosphodiesterase variants by elevated cAMP content in human myometrial cells in culture. Endocrinology. 1999;140:3228–3237. [PubMed]
  • Nilius B, Droogmans G. Ion channels and their functional role in vascular endothelium. Physiol Rev. 2001;81:1415–1459. [PubMed]
  • Oike M, Droogmans G, Nilius B. Amplitude modulation of Ca2+ signals induced by histamine in human endothelial cells. Biochim Biophys Acta. 1994;1222:287–291. [PubMed]
  • Okruhlicova L, Tribulova N, Eckly A, Lugnier C, Slezak J. Cytochemical distribution of cyclic AMP-dependent 3′,5′-nucleotide phosphodiesterase in the rat myocardium. Histochem J. 1996;28:165–172. [PubMed]
  • Rose RJ, Liu H, Palmer D, Maurice DH. Cyclic AMP-mediated regulation of vascular smooth muscle cell cyclic AMP phosphodiesterase activity. Br J Pharmacol. 1997;122:233–240. [PubMed]
  • Schwabe U, Miyake M, Ohga Y, Daly JW. 4-(3-Cyclopentyloxy-4-methoxyphenyl)-2-pyrrolidone (ZK 62711): a potent inhibitor of adenosine cyclic 3′, 5′-monophosphate phosphodiesterases in homogenates and tissue slices from rat brain. Mol Pharmacol. 1976;12:900–910. [PubMed]
  • Sette C, Iona S, Conti M. The short-term activation of a rolipram-sensitive, cAMP-specific phosphodiesterase by thyroid-stimulating hormone in thyroid FRTL-5 cells is mediated by a cAMP-dependent phosphorylation. J Biol Chem. 1994;269:9245–9252. [PubMed]
  • Socorro L, Alexander RW, Griendling KK. Cholera toxin modulation of angiotensin II-stimulated inositol phosphate production in cultured vascular smooth muscle cells. Biochem J. 1990;265:799–807. [PubMed]
  • Swinnen JV, Joseph DR, Conti M. Molecular cloning of rat homologues of the Drosophila melanogaster dunce cAMP phosphodiesterase: evidence for a family of genes. Proc Natl Acad Sci USA. 1989;86:5325–5329.
  • Tilley DG, Maurice DH. Vascular smooth muscle cell phosphodiesterase (PDE) 3 and PDE4 activities and levels are regulated by cyclic AMP in vivo. Mol Pharmacol. 2002;62:497–506. [PubMed]
  • Tiruppathi C, Ahmmed GU, Vogel SM, Malik AB. Ca2+ signaling, TRP channels, and endothelial permeability. Microcirculation. 2006;13:693–708. [PubMed]
  • Vischer UM, Wollheim CB. Purine nucleotides induce regulated secretion of von Willebrand factor: involvement of cytosolic Ca2+ and cyclic adenosine monophosphate- dependent signaling in endothelial exocytosis. Blood. 1998;91:118–127. [PubMed]
  • von der Weid PY, Serebryakov VN, Orallo F, Bergmann C, Snetkov VA, Takeda K. Effects of ATP on cultured smooth muscle cells from rat aorta. Br J Pharmacol. 1993;108:638–645. [PubMed]
  • Zhu B, Kelly J, Vemavarapu L, Thompson WJ, Strada SJ. Activation and induction of cyclic AMP phosphodiesterase (PDE4) in rat pulmonary microvascular endothelial cells. Biochem Pharmacol. 2004;68:479–491. [PubMed]
Articles from British Journal of Pharmacology are provided here courtesy of
The British Pharmacological Society